1 / 51

A two-part lecture. First: Some Monitor Examples Then: Discussion of Scheduling

A two-part lecture. First: Some Monitor Examples Then: Discussion of Scheduling. Ken Birman. Review: Monitors. Why review something we just covered? People find it hard to get into the “concurrency mindset”

zohar
Télécharger la présentation

A two-part lecture. First: Some Monitor Examples Then: Discussion of Scheduling

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. A two-part lecture.First: Some Monitor ExamplesThen: Discussion of Scheduling Ken Birman

  2. Review: Monitors • Why review something we just covered? • People find it hard to get into the “concurrency mindset” • It takes time to get used to this style of coding, and to thinking about the correctness of concurrent objects • So: review a little, then discuss a related topic, and then (next time) tackle a new concurrency issue • Yes, this isn’t “linear”… • … but avoids “too much, too fast” … Huh?

  3. Producer Consumer using Monitors     public Object get()  { Object obj;        synchronized(not_empty) { while(n == 0)not_empty.wait();        object obj = buf[tail%N];        tail++;synchronized(this) { n--; } }        synchronized(not_full) { not_full.notify(); }        return obj;    }} public class Producer_Consumer {int N;    Object[]  buf;int n = 0, tail = 0, head = 0;    Object not_empty = new Object(); Object not_full = new Object();     public Producer_Consumer(intlen) {buf = new object[len];        N = len;    }    public void put(Object obj) {        synchronized(not_full) { while(n == N)not_full.wait(); buf[head%N] = obj;        head++;synchronized(this) { n++; } }        synchronized(not_empty) { not_empty.notify(); }    }

  4. Subtle aspects • When updating a variable shared by producer and consumer, the code needs to lock it • Hence synchronized(this) { n++; } • But we can read the variable without interlocks: while(n == N)not_full.wait();

  5. Subtle aspects • To call wait or notify, must first call synchronized on the object that you will wait or notify on, hence • synchronized(xyz) { xyz.wait(); } • synchronized(xyz) { xyz.notify(); } • … if our code had been synchronized on something else, you get a “thread isn’t the owner of this object” exception • When you do call wait, do it inside a while loop and recheck the condition when you wake up • while(n == N) NotFull.wait(); • This way, if someone wakes up the thread (because n < N) but some other thread sneaks in and now n==N again, your code will be safe.

  6. Subtle aspects • Notice that when a thread calls wait(), if it blocks it also automatically releases the lock on the object on which wait was done • This is an elegant solution to an issue seen with semaphores • Caller did mutex.acquire()…. But now needs to call not_empty.aquire()… and this second call might block • So we need to call mutex.release()… • Danger is that some other thread “slips in” between the two • A race condition!

  7. Subtle aspects • But…. Java has a bug… • … nested synchronized() calls are “risky” • If the inner block calls wait, the outer lock won’t be automatically released • Can easily result in deadlocks • This is why our bounded buffer code synchronized on Not_Full… • … although sometimes we can even take advantage of this annoying behavior

  8. Readers and Writers public class ReadersNWriters {intNReaders = 0, NWriters = 0;    Object CanBegin = new Object();    public synchronized void BeginWrite()    { synchronized(CanBegin) {         if(NWriters == 1 || NReaders > 0)CanBegin.Wait(); NWriters = 1; }    }    public void EndWrite()    { synchronized(CanBegin) { NWriters = 0;CanBegin.Notify(); }     } public synchronized void BeginRead() {        synchronized(CanBegin) { if(NWriters == 1)CanBegin.Wait();            ++NReaders; }     }    public void EndRead()    {        synchronized(CanBegin) { if(--NReaders == 0)CanBegin.Notify(); }    }}

  9. Readers and Writers public class ReadersNWriters {intNReaders = 0, NWriters = 0;     Object CanBegin = new Object();    public synchronized voidBeginWrite()    { synchronized(CanBegin) {         if(NWriters == 1 || NReaders > 0)CanBegin.Wait(); NWriters = 1; }    }    public void EndWrite()    { synchronized(CanBegin) { NWriters = 0;CanBegin.Notify(); }     } public synchronized void BeginRead() {        synchronized(CanBegin) { if(NWriters == 1)CanBegin.Wait();            ++NReaders; }     }    public void EndRead()    {        synchronized(CanBegin) { if(--NReaders== 0)CanBegin.Notify(); }    }}

  10. Readers and Writers public class ReadersNWriters {intNReaders = 0, NWriters = 0;     Object CanBegin = new Object();    public synchronized void BeginWrite()    { synchronized(CanBegin) {         if(NWriters == 1 || NReaders > 0)CanBegin.Wait(); NWriters = 1; }    }    public void EndWrite()    { synchronized(CanBegin) { NWriters = 0;CanBegin.Notify(); }     } public synchronized void BeginRead() {        synchronized(CanBegin) { if(NWriters == 1)CanBegin.Wait();            ++NReaders; }     }    public void EndRead()    {        synchronized(CanBegin) { if(--NReaders == 0)CanBegin.Notify(); }    }}

  11. Understanding the Solution • If any thread is waiting, no other thread gets into BeginRead or BeginWrite • This is because of the “bug” mentioned before • When we nest synchronization blocks, a wait will only release the lock on the object we wait on… not the outer synchronization lock

  12. Readers and Writers public class ReadersNWriters {intNReaders = 0, NWriters = 0;     Object CanBegin = new Object();public synchronized void BeginWrite()    { synchronized(CanBegin) {         if(NWriters == 1 || NReaders > 0)CanBegin.Wait(); NWriters = 1; }    }    public void EndWrite()    { synchronized(CanBegin) { NWriters = 0;CanBegin.Notify(); }     } public synchronized void BeginRead() {synchronized(CanBegin) { if(NWriters == 1)CanBegin.Wait();            ++NReaders; }     }    public void EndRead()    {        synchronized(CanBegin) { if(--NReaders == 0)CanBegin.Notify(); }    }}

  13. Understanding the Solution • A writer can enter if there are no other active writers and no readers are waiting

  14. Readers and Writers public class ReadersNWriters {intNReaders = 0, NWriters = 0;     Object CanBegin = new Object();    public synchronized void BeginWrite()    { synchronized(CanBegin) { if(NWriters == 1 || NReaders > 0)CanBegin.Wait(); NWriters = 1; }    }    public void EndWrite()    { synchronized(CanBegin) { NWriters = 0;CanBegin.Notify(); }     } public synchronized void BeginRead() {        synchronized(CanBegin) { if(NWriters == 1)CanBegin.Wait();            ++NReaders; }     }    public void EndRead()    {        synchronized(CanBegin) { if(--NReaders == 0)CanBegin.Notify(); }    }}

  15. Understanding the Solution • A reader can enter if • There are no writers active or waiting • So we can have many readers active all at once • Otherwise, a reader waits (maybe many do)

  16. Readers and Writers public class ReadersNWriters {intNReaders = 0, NWriters = 0;     Object CanBegin = new Object();    public synchronized void BeginWrite()    { synchronized(CanBegin) {         if(NWriters == 1 || NReaders > 0)CanBegin.Wait(); NWriters = 1; }    }    public void EndWrite()    { synchronized(CanBegin) { NWriters = 0;CanBegin.Notify(); }     } public synchronized void BeginRead() {        synchronized(CanBegin) { if(NWriters == 1)CanBegin.Wait();            ++NReaders; }     }    public void EndRead()    {        synchronized(CanBegin) { if(--NReaders == 0)CanBegin.Notify(); }    }}

  17. Understanding the Solution • When a writer finishes, it lets one thread in, if someone is waiting • Doesn’t matter if this is one writer, or one reader • If a reader gets in, and then another shows up, it will get in too… once it gets past the “entry” synchronization • Similarly, when a reader finishes, if it was the last reader, it lets a writer in (if any is there)

  18. Readers and Writers public class ReadersNWriters {intNReaders = 0, NWriters = 0;     Object CanBegin = new Object();    public synchronized void BeginWrite()    { synchronized(CanBegin) {         if(NWriters == 1 || NReaders > 0)CanBegin.Wait(); NWriters = 1; }    }    public void EndWrite()    { synchronized(CanBegin) { NWriters = 0;CanBegin.Notify(); }     } public synchronized void BeginRead() {        synchronized(CanBegin) { if(NWriters == 1)CanBegin.Wait();            ++NReaders; }     }    public void EndRead()    {        synchronized(CanBegin) { if(--NReaders == 0)CanBegin.Notify(); }    }}

  19. Understanding the Solution • It wants to be fair • If a writer is waiting, readers queue up “outside” • If a reader (or another writer) is active or waiting, writers queue up

  20. Comparison with Semaphores • With semaphores we never had a “fair” solution • In fact it can be done, in much the same way • The monitor is relatively simple… and it works! • In general, monitors are relatively less error prone • Still, this particular one isn’t so easy to derive or understand, because it makes subtle use of the synchronization lock • Our advice: teams should agree to use monitor style

  21. Slight topic shift… Scheduling

  22. Why discuss? • Hidden under the surface in our work up to now is the role of the scheduler • It watches the pool of threads, or processes • And when the current thread/process waits, or has been running for too long • If necessary, “preempts” the current thread/process • Selects something else to run • Context switches to it

  23. OK… so who cares? • We’ve worried about the fairness of our solutions • But implicit in this is the assumption that the scheduler itself is reasonably fair • At a minimum, that every runnable thread/process gets plenty of opportunities to run • Let’s look more closely at how schedulers really work

  24. Process Scheduling • Rest of lecture: “process” and “thread” used interchangeably • Many processes in “ready” state • Which ready process to pick to run on the CPU? • 0 ready processes: run idle loop • 1 ready process: easy! But if > 1 ready process: what to do? New Ready Running Exit Waiting

  25. Some schedulers have no choice • For example, in Java, if you “notify” a thread, it will be the next thread to obtain the synchronization lock • … even if other threads are waiting • But often, there are situations with • Multiple active processes, or threads • All want to run…

  26. When does scheduler run? • Non-preemptive minimum • Process runs until voluntarily relinquish CPU • process blocks on an event (e.g., I/O or synchronization) • process terminates • Preemptive minimum • All of the above, plus: • Event completes: process moves from blocked to ready • Timer interrupts • Implementation: process can be interrupted in favor of another Exit New Running Ready Waiting

  27. emacs Editor Process Model • Process alternates between CPU and I/O bursts • CPU-bound jobs: Long CPU bursts • I/O-bound: Short CPU bursts • I/O burst = process idle, switch to another “for free” • Problem: don’t know job’s type before running • An underlying assumption: • “response time” most important for interactive jobs (I/O bound) Matrix multiply

  28. Scheduling Evaluation Metrics • Many quantitative criteria for evaluating schedalgo: • CPU utilization: percentage of time the CPU is not idle • Throughput: completed processes per time unit • Turnaround time: submission to completion • Waiting time: time spent on the ready queue • Response time: response latency • Predictability: variance in any of these measures • The right metric depends on the context

  29. “The perfect CPU scheduler” • Minimize latency: response or job completion time • Maximize throughput: Maximize jobs / time. • Maximize utilization: keep I/O devices busy. • Recurring theme with OS scheduling • Fairness: everyone makes progress, no one starves

  30. Problem Cases • Blindness about job types • I/O goes idle • Optimization involves favoring jobs of type “A” over “B”. • Lots of A’s? B’s starve • Interactive process trapped behind others. • Response time sucks for no reason • Priorities: A depends on B. A’s priority > B’s. • B never runs

  31. Scheduling Algorithms FCFS • First-come First-served (FCFS) (FIFO) • Jobs are scheduled in order of arrival • Non-preemptive • Problem: • Average waiting time depends on arrival order • Advantage: really simple! P1 P2 P3 time 0 16 20 24 P2 P3 P1 0 4 8 24

  32. Convoy Effect • A CPU bound job will hold CPU until done, • or it causes an I/O burst • rare occurrence, since the thread is CPU-bound  long periods where no I/O requests issued, and CPU held • Result: poor I/O device utilization • Example: one CPU bound job, many I/O bound • CPU bound runs (I/O devices idle) • CPU bound blocks • I/O bound job(s) run, quickly block on I/O • CPU bound runs again • I/O completes • CPU bound still runs while I/O devices idle (continues…) • Simple hack: run process whose I/O completed? • What is a potential problem?

  33. Scheduling Algorithms LIFO • Last-In First-out (LIFO) • Newly arrived jobs are placed at head of ready queue • Improves response time for newly created threads • Problem: • May lead to starvation – early processes may never get CPU

  34. Problem • You work as a short-order cook • Customers come in and specify which dish they want • Each dish takes a different amount of time to prepare • Your goal: • minimize average time the customers wait for their food • What strategy would you use ? • Note: most restaurants use FCFS.

  35. Scheduling Algorithms: SJF • Shortest Job First (SJF) • Choose the job with the shortest next CPU burst • Provably optimal for minimizing average waiting time • Problem: • Impossible to know the length of the next CPU burst P1 P3 P2 0 15 21 24 P2 P2 P3 P3 P1 0 3 9 24

  36. Scheduling Algorithms SRTF • SJF can be either preemptive or non-preemptive • New, short job arrives; current process has long time to execute • Preemptive SJF is called shortest remaining time first P2 P3 P1 0 21 6 10 P3 P1 P2 P1 0 10 13 24 6

  37. Shortest Job First Prediction • Approximate next CPU-burst duration • from the durations of the previous bursts • The past can be a good predictor of the future • No need to remember entire past history • Useexponential average: tnduration of the nth CPU burst n+1predicted duration of the (n+1)st CPU burst n+1 = tn + (1- ) n where 0    1  determines the weight placed on past behavior

  38. Priority Scheduling • Priority Scheduling • Choose next job based on priority • For SJF, priority = expected CPU burst • Can be either preemptive or non-preemptive • Problem: • Starvation: jobs can wait indefinitely • Solution to starvation • Age processes: increase priority as a function of waiting time

  39. Round Robin • Round Robin (RR) • Often used for timesharing • Ready queue is treated as a circular queue (FIFO) • Each process is given a time slice called a quantum • It is run for the quantum or until it blocks • RR allocates the CPU uniformly (fairly) across participants. • If average queue length is n, each participant gets 1/n

  40. P1 P2 P3 P4 P1 P3 P4 P1 P3 P3 RR with Time Quantum = 20 ProcessBurst Time P1 53 P2 17 P3 68 P4 24 • The Gantt chart is: • Higher average turnaround than SJF, • But better response time 0 20 37 57 77 97 117 121 134 154 162

  41. Turnaround Time w/ Time Quanta

  42. RR: Choice of Time Quantum • Performance depends on length of the timeslice • Context switching isn’t a free operation. • If timeslice time is set too high • attempting to amortize context switch cost, you get FCFS. • i.e. processes will finish or block before their slice is up anyway • If it’s set too low • you’re spending all of your time context switching between threads. • Timeslice frequently set to ~100 milliseconds • Context switches typically cost < 1 millisecond Moral: Context switch is usually negligible (< 1% per timeslice) unless you context switch too frequently and lose all productivity

  43. Scheduling Algorithms • Multi-level Queue Scheduling • Implement multiple ready queues based on job “type” • interactive processes • CPU-bound processes • batch jobs • system processes • student programs • Different queues may be scheduled using different algos • Intra-queue CPU allocation is either strict or proportional • Problem: Classifying jobs into queues is difficult • A process may have CPU-bound phases as well as interactive ones

  44. Multilevel Queue Scheduling Highest priority System Processes Interactive Processes Batch Processes Student Processes Lowest priority

  45. SchedulingAlgorithms • Multi-level Feedback Queues • Implement multiple ready queues • Different queues may be scheduled using different algorithms • Just like multilevel queue scheduling, but assignments are not static • Jobs move from queue to queue based on feedback • Feedback = The behavior of the job, • e.g. does it require the full quantum for computation, or • does it perform frequent I/O ? • Very general algorithm • Need to select parameters for: • Number of queues • Scheduling algorithm within each queue • When to upgrade and downgrade a job

  46. Multilevel Feedback Queues Highest priority Quantum = 2 Quantum = 4 Quantum = 8 FCFS Lowest priority

  47. A Multi-level System I/O bound jobs high priority CPU bound jobs high low timeslice

  48. Thread Scheduling Since all threads share code & data segments • Option 1: Ignore this fact • Option 2: Gang scheduling • run all threads belonging to a process together (multiprocessor only) • if a thread needs to synchronize with another thread • the other one is available and active • Option 3: Two-level scheduling: • Medium level scheduler • schedule processes, and within each process, schedule threads • reduce context switching overhead and improve cache hit ratio • Option 4: Space-based affinity: • assign threads to processors (multiprocessor only) • improve cache hit ratio, but can bite under low-load condition

  49. Real-time Scheduling • Real-time processes have timing constraints • Expressed as deadlines or rate requirements • Common RT scheduling policies • Rate monotonic • Just one scalar priority related to the periodicity of the job • Priority = 1/rate • Static • Earliest deadline first (EDF) • Dynamic but more complex • Priority = deadline • Both require admission control to provide guarantees

  50. Postscript • The best schemes are adaptive. • To do absolutely best we’d have to predict the future. • Most current algorithms give highest priority to those that need the least! • Scheduling become increasingly ad hoc over the years. • 1960s papers very math heavy, now mostly “tweak and see”

More Related